Optimized capacity allocation in a UAM vertiport network utilizing efficient ride matching

Abstract

In this study, we present an approach that allows to design efficient UAM fleets and corresponding vertiport networks for specific demand patterns. Therefore, we apply a trajectory-based simulation model that controls the circulation of vehicles in a UAM network. The model allocates vehicles from an unlimited fleet pool to the requested missions, such that the boundary conditions of the optimization are fulfilled. Applying a combination of graph-based optimization and solving integer linear programming problems, a ride matching algorithm is implemented that minimizes empty relocation flights in the network and reduces the fleet size. The results comprise the quantification of a fleet pool, defined by fleet size, fleet mix, and starting positions in the vertiport network. We analyze a set of 20 vertiports in the City of Hamburg, Germany, regarding the local peak loads of parking positions that are needed for battery charging and waiting periods of unoccupied vehicles. The results show that the reduction of battery charging time has a significant impact on fleet size, which affects the minimum ground infrastructure requirements as well. Finally, the fleet analysis shows that average load factors of 45% are feasible at fleet sizes with varying occupancy rates of up to 80%.